RÉPONSE À UNE SUPPLÉMENTATION EN ACIDES GRAS
OMÉGA-3 D’ORIGINE MARINE
Novel Genetic Loci Associated with the Plasma Triglyceride Response to an Omega-3 Fatty Acid Supplementation
Bastien Vallée Marcotte, Hubert Cormier, Frédéric Guénard, Iwona Rudkowska, Simone Lemieux, Patrick Couture, Marie-Claude Vohl
Vallee Marcotte B, Cormier H, Guenard F, Rudkowska I, Lemieux S, Couture P, Vohl MC. Novel genetic loci associated with the plasma triglyceride response to an omega-3 fatty acid supplementation. J Nutrigenet Nutrigenomics 2016;9:1-11.
Cet article a été publié dans le Journal of Nutrigenetics and Nutrigenomics, en mai 2016 [142]. Mon rôle dans cette étude a d’abord été de participer au génotypage en laboratoire, d’effectuer les analyses statistiques et de rédiger le papier. Ces résultats ont été présentés oralement au congrès de la Société québécoise de lipidologie (SQLNM), en février 2016, et au Symposium des étudiants de l'institut sur la nutrition et les aliments fonctionnels (INAF) en avril 2016. Ils ont également été présentés par affiche à la réunion annuelle de la Société canadienne de Nutrition (CNS) en mai 2016.
29
R
ÉSUMÉ ObjectifVérifier si des SNPs à l’intérieur des gènes IQCJ, NXPH1, PHF17 et MYB sont associés à la réponse des TG plasmatiques suite à une supplémentation en AG n-3 d’origine marine.
Méthodes
208 sujets ont été supplémentés en AG n-3 à raison de 3g d’AG n-3 par jour (1,9-2,2g d’AEP et 1,1g d’ADH) pendant six semaines. Les lipides plasmatiques étaient mesurés avant et après la supplémentation. 67 SNPs ont été sélectionnés afin d’augmenter la densité de marqueurs près des signaux significatifs du GWAS.
Résultats
Dans un modèle à mesures répétées, des effets du génotype et des interactions gène-diète affectaient les niveaux de TG plasmatiques suite à la supplémentation. Des effets du génotype ont été observés avec deux SNPs de NXPH1 et des interactions gène-diète ont été observées avec 10 SNPs d’IQCJ, quatre de NXPH1 et trois de MYB. Les répondeurs positifs et négatifs à la supplémentation avaient des fréquences génotypiques différentes pour neuf SNPs d’IQCJ, deux de NXPH1 et deux de MYB.
Conclusion
La cartographie fine des loci associés au GWAS a permis d’identifier des SNPS qui expliquent en partie l’importante variabilité individuelle de la réponse des TG à une supplémentation en AG n-3 d’origine marine.
30
NOVEL GENETIC LOCI ASSOCIATED WITH THE PLASMA TRIGLYCERIDE RESPONSE TO AN OMEGA-3 FATTY ACID SUPPLEMENTATION
Institute of Nutrition and Functional Foods (INAF)
Bastien Vallée Marcotte1, Hubert Cormier1, Frédéric Guénard1, Iwona Rudkowska2, Simone Lemieux1, Patrick Couture 1,2, Marie-Claude Vohl1,2*
1. Institute of Nutrition and Functional Foods (INAF), Laval University, 2440 Hochelaga Blvd, Quebec, QC, Canada
2. CHU de Québec Research Center – Endocrinology and Nephrology, 2705 Laurier Blvd, Quebec, QC, Canada
Corresponding author*: Marie-Claude Vohl Ph.D.
Institute of Nutrition and Functional Foods (INAF) 2440 Hochelaga Blvd.
Quebec, QC, Canada G1V 0A6
Tel.: (418) 656-2131 ext. 4676, Fax: (418) 656-5877 E-Mail: marie-claude.vohl@fsaa.ulaval.ca Email addresses: BVM: bastien.vallee-marcotte.1@ulaval.ca HC: hubert.cormier.1@ulaval.ca FG: frederic.guenard@fsaa.ulaval.ca IR: iwona.rudkowska@crchudequebec.ulaval.ca SL: simone.lemieux@fsaa.ulaval.ca PC: patrick.couture@crchudequebec.ulaval.ca MCV: marie-claude.vohl@fsaa.ulaval.ca
31 ABSTRACT
Background
A recent genome-wide association study (GWAS) by our group identified 13 loci associated with the plasma triglyceride (TG) response to omega-3 (n-3) fatty acid (FA) supplementation. This study aimed to test whether single nucleotide polymorphisms (SNPs) within the IQCJ,
NXPH1, PHF17 and MYB genes are associated with the plasma TG response to an n-3 FA
supplementation.
Methods
208 subjects followed a 6-week n-3 FA supplementation of 5g/d of fish oil (1.9–2.2g of EPA and 1.1g of DHA). Measurements of plasma lipids were made before and after the supplementation. 67 tagged SNPs were selected to increase the density of markers near GWAS hits.
Results
In a repeated model, independent effects of the genotype and the gene by supplementation interaction were associated with plasma TG. Genotype effects were observed with two SNPs of NXPH1 and gene-diet interactions were observed with ten SNPs of IQCJ, four of NXPH1 and three of MYB. Positive and negative responders showed different genotype frequencies with nine SNPs of IQCJ, two of NXPH1 and two of MYB.
Conclusion
Fine mapping in GWAS-associated loci allowed the identification of SNPs explaining partly the large inter-individual variability observed in plasma TG levels in response to an n-3 FA supplementation.
Keywords
Gene-diet interactions, plasma lipid levels, omega-3 fatty acids, genome-wide association study, nutrigenetics.
32 BACKGROUND
It has been demonstrated that omega-3 (n-3) fatty acids (FAs) from marine sources, including eicosapentaenoic acid (EPA) and docosahexaenoic acid (DHA), are beneficial in cardiovascular disease (CVD) prevention [1]. More specifically, the consumption of n-3 FA is associated with a decrease in plasma triglyceride (TG) levels, which are recognized to be a CVD risk factor [2]. However, there is an important and well-recognized interindividual variability in the plasma TG response to an n-3 FA supplementation. Our research group has previously reported that 29% of all participants of the Fatty Acid Sensor (FAS) study, who received a 6-week n-3 FA supplementation of 5 g of fish oil (1.9--2.2 g EPA and 1.1 g DHA) per day, did not reduce their plasma TG levels [3]. Likewise, in the FINGEN study, 31% of all volunteers did not show any reduction of their plasma TG levels after an 8-week supplementation of 1.8 g of n-3 FA per day [4]. This heterogeneity in the plasma TG response may be partly due to genetic factors [5]. Numerous research groups around the world have been interested in the study of associations between single-nucleotide polymorphisms (SNPs) in candidate genes such as apolipoprotein E (APOE) or the peroxisome proliferator-activated receptors (PPARs) and their effect on plasma TG levels in response to n-3 FA [4, 6--9]. However, these variations account for a very small proportion of the variance in the plasma TG levels in response to an n-3 FA supplementation. Other not yet identified genetic variations are thus likely to contribute to the heterogeneity of the response. Genome-wide association studies (GWAS) provide a more complete and nonrestrictive approach to test the impact of SNPs on a phenotype.
Recently, a GWAS performed by our research group on subjects of the FAS study identified 13 SNPs associated with plasma TG levels [10]. A genetic risk score constructed with these loci explained 21.53% of the variability in plasma TG levels in response to an n-3 FA supplementation. Most of these SNPs were located within or in the vicinity of the IQ motif- containing J (IQCJ), neurexophilin-1 (NXPH1), PHD finger protein 17 (PHF17) and V-MYB avian myeloblastosis viral oncogene homolog (MYB) genes. IQCJ is mostly expressed in the brain and is part of an isoform of a transcriptional unit named IQCJ-SCHIP1 that binds two distinct genes, IQCJ and SCHIP1, encoding different proteins. IQCJ-SCHIP1 is implicated in the neuronal development, particularly of the nodes of Ranvier and of the axon initial segment [11, 12]. NXPH1 encodes the neurexophilin-1 protein, which binds to α-neurexins in the synapse and is implicated in neuronal regulation [13--15]. PHF17 encodes the Jade-1 protein, which plays a role in cell growth, apoptosis and cancer [16--18]. Jade-1 is stabilized by the von
33
Hippel-Lindau tumor suppressor [16]. It also interacts with HBO1 histone acetyltransferase (HAT) to promote acetylation of nucleosomal histones and to increase cell proliferation [18, 19]. MYB encodes a transcriptional factor whose role is mainly to regularize hematopoiesis, tumorigenesis and cell growth [20--22].
Thus, the objective of the present study was to increase the density of markers within the IQCJ,
NXPH1, PHF17 and MYB genes and to test their association with plasma TG levels in response
to an n-3 FA supplementation. The hypothesis was that genetic factors exert influence on plasma TG levels in response to an n-3 FA supplementation.
34 SUBJECTS AND METHODS
Study Population
A total of 254 subjects from the Quebec City metropolitan area were recruited between September 2009 and December 2011 to participate in the FAS study. Recruitment was made via announcements in local newspapers and electronic messages sent to university students and employees. To be eligible, participants had to be aged between 18 and 50 years and to have a body mass index (BMI) between 25 and 40. They also had to be nonsmokers and be free of any thyroid or metabolic disorders requiring treatment, such as diabetes, hypertension, dyslipidemia or coronary heart disease. Participants were excluded from the study if they had taken n-3 FA supplements for at least 6 months prior to the beginning of the study. In sum, 210 participants completed the intervention protocol. Two of them did not have plasma TG level data available for further analyses, leaving 208 subjects to be included in the study sample. Statistical analyses were therefore conducted on 208 subjects. Subjects who completed the intervention protocol were divided into two subgroups: positive and negative responders to the n-3 FA supplementation. This classification was made according to the change in their plasma TG levels. Participants who had ΔTG ≥0 were considered as negative responders and those whose ΔTG was <0 were considered as positive responders.
Study Design and Diets
The whole study design and diets have been previously reported [10]. Briefly, subjects followed a run-in period of 2 weeks. During these 2 weeks, they received individual dietary instructions by a trained registered dietitian to achieve the recommendations from Canada’s Food Guide to Healthy Eating. Recommendations were given to ensure constant n-3 FA dietary intake and body weight stability throughout the protocol. After the run-in period, the participants received a bottle containing the needed n-3 FA capsules (Ocean Nutrition, Dartmouth, N.S., Canada) that would be taken for the next 6 weeks. Each capsule contained 1 g of fish oil concentrate. They had to take 5 capsules a day, providing 3 g of n-3 FA, including 1.9–2.2 g of EPA and 1.1 g of DHA. Furthermore, they were asked to report any deviation from the protocol or experienced side effects and to write down their alcohol intake as well as their fish consumption. They also received detailed oral and written instructions on their diet before each phase.
Laboratory Methods Plasma Lipids
35
Methods to measure plasma lipids have previously been detailed [10]. Briefly, blood samples were collected after a 12-hour overnight fast and 48-hour alcohol abstinence. Blood samples were taken before the run-in period to verify whether individuals were presenting any metabolic disorders. The remaining participants who were eligible to the study had blood samples taken prior to and after the n-3 FA supplementation period. Enzymatic assays were used to measure plasma total cholesterol and TG concentrations [23, 24].
SNP Selection and Genotyping
SNPs in IQCJ, NXPH1, PHF17 and MYB were identified via the International HapMap Project SNP database, based on the National Center for Biotechnology Information (NCBI) B36 assembly Data Rel 28, phase II + III, build 126. The Gene Tagger procedure in the Haploview software v4.2 was used to determine tagged SNPs with a minor allele frequency (MAF) >5% and pairwise tagging (r2 ≥ 0.80) located in gene regions and surrounding regions (2 kb). The mean r2 was 0.96 for IQCJ, 0.96 for NXPH1, 0.97 for PHF17 and 0.95 for MYB. Sixteen tag SNPs in IQCJ, 34 in NXPH1, eight in PHF17 and nine in MYB were selected, so that ≥85% of all common variations (MAF >5%) in these genes were covered. The GenElute Gel Extraction Kit (Sigma-Aldrich Co., St. Louis, Mo., USA) was used to extract genomic DNA from the blood samples. Selected SNPs were genotyped using TaqMan technology. More specifically, 2.5 µl of each genomic DNA (40 ng/µl) and 2.5 µl of OpenArray® Genotyping Master Mix (Life Technologies, Carlsbad, Calif., USA) were mixed with a validated primer in a 384-well plate and loaded onto the genotyping plates with the QuantStudioTM 12K Flex OpenArray® AccuFillTM System (Life Technologies). Genotyping was conducted on the QuantStudioTM 12K Flex Real-Time PCR System (Life Technologies). Thereafter, results were analyzed in TaqMan Genotyper v1.3 (Life Technologies).
Statistical Analyses
All statistical analyses were run in SAS Statistical Software v9.2 (SAS Institute, Cary, N.C., USA), except for the ALLELE procedure, which was conducted using SAS Genetics v9.3. This procedure was used to test genotype distributions for any deviation from the Hardy-Weinberg equilibrium and to calculate the MAF. Normal distribution was evaluated with the box plot, skewness and kurtosis ranges. Abnormally distributed variables were log10 transformed. Rare genotype homozygotes presenting a genotype frequency <5% were merged with heterozygotes for statistical analyses (dominant model). Otherwise, all three genotypes were analyzed separately as three groups, namely major allele homozygotes, heterozygotes and minor allele
36
homozygotes (additive model). The MIXED procedure for repeated measures was used to test whether there was an effect of genotype, supplementation and genotype-supplementation (gene-diet) interaction on TG levels after the n-3 FA supplementation in a model adjusted for age, sex and BMI. Multiple testing was then performed using the MULTTEST procedure to control for the false discovery rate. To verify whether dominant SNPs drive the associations with TG levels, a stepwise regression was conducted for each gene independently using the REG procedure. The MIXED procedure for repeated measures was also used to test for significant differences in the study sample before and after supplementation in a model adjusted for age, sex and BMI. Finally, the FREQ procedure was used to verify differences in genotype frequency distribution between positive and negative responders to the n-3 FA supplementation. p ≤ 0.05 was defined as the statistical significance threshold.
Statement of Ethics
The study was approved by the Laval University and CHU de Québec ethics committees and was performed in accordance with the principles of the Declaration of Helsinki. All participants provided written, informed consent.
37 RESULTS
Table 1 shows allele frequencies of selected SNPs. All SNPs were in Hardy-Weinberg equilibrium and were therefore kept for further analyses. The percent coverage of each gene was 87% for IQCJ, 85% for NXPH1, 96% for PHF17 and 100% for MYB. The vast majority of SNPs were located within introns. Three SNPs of PHF17, namely rs2217023, rs13143771 and rs13142964, were located in the upstream region of the gene, and rs13148510 was located in the 3′ UTR region. Finally, one SNP of MYB (rs210936) was located in the downstream region of the gene.
Baseline characteristics of the subjects are shown in table 2. The average BMI before and after supplementation of all participants was >25 as specified by the study design. The average plasma TG levels before the supplementation were above the cut-point value of 1.129 mmol/l according to the American Heart Association’s recommendation for optimal plasma TG levels [2].
For each selected SNP, we tested the independent effect of genotype, supplementation (time) and genotype-supplementation (interaction) on plasma TG levels (table 3; for changes in n-3 levels after supplementation, see online suppl. table 1; for all online suppl. material, see www.karger.com/doi/10.1159/000446024). First, as expected, the supplementation alone had a significant effect on plasma TG levels (p < 0.006, for all). Second, two SNPs of NXPH1 (rs2107779 and rs2107474) were associated with TG levels (p < 0.05, for both) during the supplementation. Two SNPs of NXPH1 and another of MYB had marginal but not significant effects on TG levels (p < 0.1, for all). Third, significant gene-diet interaction effects were observed with ten SNPs of IQCJ (rs2044704, rs1962071, rs6800211, rs17782879, rs1868414, rs2595260, rs9827242, rs1449009, rs2621309 and rs61332355), four SNPs of NXPH1 (rs7806226, rs7805772, rs2349780 and rs6974252) and three SNPs of MYB (rs9321493, rs11154794 and rs210962) (p < 0.05, for all). Marginal effects of gene-diet interactions were observed with two additional SNPs, located in IQCJ and NXPH1, respectively (p < 0.1, for both). Neither genotype effects nor gene-diet interaction effects were found with SNPs of the
PHF17 gene. Overall, four SNPs remained significant after multiple testing with the false
discovery rate, namely rs1449009 (p = 0.03), rs2621309 (p = 0.03) and rs61332355 (p = 0.01) in IQCJ and rs7805772 (p = 0.03) in NXPH1. The stepwise regression revealed four dominant SNPs driving the associations with TG, namely rs61332355 (p = 0.0001) and rs9827242 (p = 0.04) in IQCJ, rs7805772 (p = 0.007) in NXPH1 and rs11154794 (p = 0.007) in MYB.
38
Significant differences were observed in the genotype distribution between positive and negative responders for several SNPs in IQCJ, NXPH1 and MYB as shown in table 4. Globally, higher frequencies of the minor allele were observed in negative responders, while for two SNPs, rs1868414 and rs9827242, there was a higher frequency of carriers of the minor allele in positive responders than in negative responders. There was no significant difference in the genotype distribution for SNPs from PHF17.
39 DISCUSSION
In the present study, we narrowed down regions previously identified in a GWAS performed on the FAS study by increasing the density of markers in IQCJ, NXPH1, PHF17 and MYB and testing for associations with plasma TG levels during the supplementation. To do so, 67 SNPs were selected in order to cover ≥85% of the genetic variability near GWAS hits. This study is the first to report on a follow-up of GWAS signals associated with plasma TG levels in a supplementation protocol with n-3 FAs.
In this study, it was observed that the n-3 FA supplementation alone induced a decrease in plasma TG levels, as previously reported [4]. A mean decrease of 0.19 mmol/l (15.7%) in TG levels was observed after supplementation compared to before supplementation for all subjects. Considering that lipid-lowering drugs such as ezetimibe, statins or both can reduce TG levels by 15--20% in patients with hypercholesterolemia according to a recent meta-analysis, these results support the notion that n-3 FA supplements can be an effective approach to treat hypertriglyceridemia in patients who respond positively to n-3 FA supplementation [25]. Furthermore, two SNPs were independently associated with plasma TG levels, and significant gene-diet interaction effects were found for 17 SNPs. Thirteen SNPs had a different genotype frequency distribution when comparing positive and negative responders. There are some similarities in the results presented in tables 3 and 4. Indeed, nine out of the ten SNPs whose genotype frequency distribution was significantly different between positive and negative responders have either been found to exert a significant influence on plasma TG levels or to show a gene-diet interaction with the supplementation.
Most of these associations were observed with SNPs in the IQCJ and NXPH1 genes. Both of these genes are mainly known to be expressed in the brain and to participate in nervous system functions. The IQCJ gene alone has been barely studied. IQCJ is bound to SCHIP1 to form a fusion gene called IQCJ-SCHIP1, and up to now, the vast majority of research has been done on the whole IQCJ-SCHIP1 segment. Nevertheless, it has been demonstrated that the IQCJ-
SCHIP1 protein contains a calmodulin-binding IQ motif at the N-terminus which SCHIP1 does
not have, thus suggesting that the latter has a different biological role [11]. SCHIP1 has also been associated with neurofibromatosis type 2, a tumor suppressor protein [11]. According to the literature, IQCJ-SCHIP1 is mostly involved in neuronal function, and its protein has been observed in the cytoplasm, actin-rich regions, differentiated PC12 cells and neurite extensions [11, 12]. As to NXPH1, it is a member of the neurexophilin family, composed of four genes
40
encoding for related glycoproteins that act as neuropeptides and interact with neurexins [13-- 15].
For two of these genes, IQCJ and NXPH1, very few associations with plasma lipid levels have been reported. It has recently been reported in the Genetics of Lipid Lowering Drugs and Diet
Network (GOLDN) study, which aimed to investigate SNPs responsible for the heterogeneity
observed in the response to a lipid-lowering therapy with fenofibrate, that these two genes may actually play a role in lipid metabolism [26]. Indeed, Kraja et al. [26] conducted a GWAS and demonstrated that one SNP in IQCJ-SCHIP1 (rs6805526) was associated with very-low- density lipoprotein particle clearance and TG levels. It is well known that a decrease in TG synthesis through fish oil rich in n-3 FAs is also associated with a decrease in very-low-density lipoprotein secretion from the liver [27]. Moreover, fenofibrate and n-3 FA have similar mechanisms of action as they can both activate PPAR-α, a nuclear receptor involved in the regulation of intra- and extracellular lipid metabolism, and they have beneficial effects on inflammation [28, 29]. Like n-3 FAs, fenofibrate therapy is used to lower plasma TG levels, and its effectiveness to reduce the risk of CVD events in patients with hypertriglyceridemia has been demonstrated [30]. However, it should be noted that rs6805526 is located in the SCHIP1 section of IQCJ-SCHIP1 and is thus not part of IQCJ alone. rs6805526 is not in linkage